GB2326956A

GB2326956A - Control means for controlling a load

Info

Publication number: GB2326956A
Application number: GB9813896A
Authority: GB
Inventors: Andreas Kellner
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 1997-07-01
Filing date: 1998-06-26
Publication date: 1999-01-06
Anticipated expiration: 2018-06-26
Also published as: GB2326956B; FR2765699B1; JP4612126B2; JPH1173233A; GB9813896D0; DE19727944A1; FR2765699A1

Description

1 2326956 METHOD AND CONTROL MEANS FOR CONTROLLING A LOAD The present

invention relates to a method and control means for controlling a load.

A method and devices for the control of loads are described in DE-OS 35 29 742, in which the load is controlled in pulsed manner with a keying ratio preset in dependence on the resistance of the load and a desired target current value. The current flowing through the load is detected and compared with the target value and a regulator varies the keying ratio in dependence on this comparison so that the target current value is achieved.

This procedure entails problems in the case of pulsed drive controls when conflicting demands are made on the system. The drive control frequency should be as low as possible for the avoidance of static friction. In the case of a lower drive control frequency, the load remains in constant movement and no static friction arises. A lower drive control frequency, however, requires the current to fluctuate at a corresponding frequency. In order to be able to make available a suitable actual value, the actual value has to be filtered by a low-pass filter. In the case of a lower cycle frequency, a long dead time arises due to the low-pass filter. This in turn has the consequence that the actual value reacts only very slowly to changed target values. The result is a dynamically slow regulator which reacts to changes in the target value only after a large delay time.

There is thus a need for a dynamically rapid regulation which may be able to react rapidly to changes in a target value in the context of control of a load with a pulsed drive control of low cycle frequency.

1 According to a first aspect of the invention there is provided a method for the control of a load, wherein the load is controlled in drive in pulsed manner at a presettable keying ratio and the keying ratio is presettable in dependence on a resistance value of the load and a target current value, characterised in that the resistance value of the load is adapted.

Preferably, the adaptation of the resistance value of the load is carried out starting from a regulating deviation between an actual current value and the target current value. The resistance value is preferably presettable starting from a preliminary control value and a correction factor, and the correction factor is preferably presettable, for example by integrating means, starting from a comparison between the actual current value and the 2 target current value. The correction factor and the preliminary control value can be multiplied and added to the preliminary control value for computation of the resistance value of the load. The actual current value is preferably filtered by a low-pass filter.

According to a second aspect of the invention there is provided control means for the control of a load, wherein the load is controlled in drive in pulsed manner at a presettable keying ratio and a control presets the keying ratio in dependence on the resistance value of the load and a target current value, characterised in that means are provided which adapt the resistance value of the load.

in the case of a method exemplifying the invention, the actual value for the current can be made to follow changes in the target value very rapidly. This is possible even with pulsed drive control and a low cycle frequency. There thus results a more rapid control behaviour by comparison with a proportionai-integral regulator with preliminary control. This also applies for low cycle frequencies and thus a long dead time of the regulated section. Moreover, the same aperiodic transient behaviour of the current-regulating loop results independently of the actual resistance of the load. No compensating resistors for temperature compensation of the load are necessary and additional power losses can thus be avoided.

An example of the method and embodiment of the control means of the present invention will now be more particularly described with reference to the accompanying drawings, in which:

Fig. 1 is a block circuit diagram of control means embodying the invention; Figs. 2a to c are diagrams showing signals, entered as a function of time, arising in the control means; and Fig. 3 is a block circuit diagram of a control stage in the control means.

Referring now to the drawings there is shown in Fig. 1 control means for a load 100, which is connected on the one hand with a supply voltage Ubat and on the other hand with ground by way of current-measuring means 110 and switching means 120.

3 A differential amplifier 130, the output signal of which is fed by way of a low-pass filter 141 to an anaiog-to-digital converter 150, is connected to both terminals of the currentmeasuring means 110. The filter 141, in the illustrated embodiment, consists of a resistor 140 arranged between the analog converter 150 and the output of the amplifier 130. The resistor 140 is connected with ground by way of a capacitor 145. The resistor 140 and the capacitor 145 form the filter 141.

The output signal of the converter 150 is applied to a control 160, which acts on the switching means 120 by a drive control signal U.

Preferably, the load 100 is a coil of an electromagnetic valve with a needle. In the case of, in particular, proportional electromagnetic valves, the current 1 flowing through the load 100 is regulated towards a preset value. A certain position of the valve arises in dependence on the current 1 flowing through the load 100.

Such proportional electromagnetic valves are used in, for example, internal combustion engines, especially for pressure regulation. Thus, for example, in a common-rail fuel supply systems the pressure in a storage device is set to a presettable value. For this purpose, a proportional electromagnetic valve is used. In dependence on the current flowing through the load 100, there is a specific fuel flow through the valve and thereby a specific pressure change.

The method is not, however, restricted to this application. It can be used for all loads in which the current must be set to a certain value.

The switching means 120 preferably has the form of a transistor, in particular a field effect transistor. The current-measuring means 110 preferably has the form of a resistive impedance. The arrangement of the eu rrent-measu ring means 110 and of the switching means 120 in Fig. 1 is merely by way of example. Equally, the current-measuring means 110 can be connected to ground or to supply voltage Ubat.

The operation of the control means is described by reference to the signals entered in the diagrams of Figs. 2a to 2c. In dependence on the desired current to flow through the load 100, the control 160 determines the drive control signal U by which the switching means 120 is controlled in drive. This signal is illustrated in Fig. 2a. The frequency or the period 4 duration T is constant in this case and preset in dependence on operational characteristic magnitudes. The period duration T is so chosen that no static friction occurs in the valve. This is avoided by the valve needle being constantly in movement. The frequency or the period duration T is preset accordingly.

As a rule, the valve represents an inductive load. This means that when the current is switched on, i.e. the voltage U has a high value, the current 1 rises. In the switched-off state, the current 1 decays.

Due to the inductance of the load, a current arises as illustrated in Fig. 2b. This signal 1 is made available by the differential amplifier 130 and is not suitable for direct further processing. Thus, as a rule, an instantaneous value is scanned in the case of digital evaluation. The value changes greatly according to scanning instant. In order to make a digital evaluation possible, the signal is filtered by means of the lowpass filter 141. The signal IF, which is illustrated in Fig. 2c, is then present at the output of the low-pass filter.

In place of the illustrated resistance-capacitance filter, other suitable filters can be used. It is a problem that a large dead time results in the actual value detection as a result of the filtering. A change in the actual current value has the consequence of a change in the signal IF only after a relatively long dead time. This dead time is, on the one hand, desired, since short-term current fluctuations as illustrated in Fig. 2b are balanced out. On the other hand, however, in the case of a change in the target value this leads to an appreciable dead time of the actual value detection.

The value of the averaged current signal IF depends substantially on the resistance R of the load and the keying ratio TV of the drive control. The keying ratio is defined by the ratio EIT between the time E, in which the switching means releases the current flow, and the period duration T.

The filtered current value IF is fed by way of the digital-to-analog converter 150 to the control 160. The control 160, starting from the signal IF, then determines the drive control signal U.

The control 160 is illustrated in more detail in Figure 3. Elements already described in Fig. 1 are denoted by corresponding reference symbols. A keying ratio presetting device 300 acts on the switching means 120 by the drive control signal U and processes the output signal R of an interlinking point 310. The output signal of a RO-presetting device 330 is present at a first input of the interlinking point 310 and at a first input of an interlinking point 320. The output signal of the point 320 is present at a second input of the point 310. The output signal K of a limiter 340, which is acted on by a signal from an integrator 350, is present at a second input of the point 320. The integrator 350 processes the output signal of an interlinking point 370, to a first input of which the output signal of the analog-to- digital converter 150 is applied. The output signal ISF of a target value filter 380 is applied to a second input of the point 370. An output signal IS of a target value presetting device 360, which also acts on the device 330 by the target value IS, is present at the input of the filter 380.

If so desired, the target value filter 380 can also be omitted.

The regulating section is marked by a dashed line. The load 100 and the low-pass filter 141 are illustrated as PT1 members.

In operation, starting from the desired target current value IS and the resistance R of the load 100, the keying ratio presetting device 300 determines the necessary keying ratio by which the switching means 120 is to be acted on. The resistive impedance R(T) of the load 100 is dependent on the temperature T. In that case, for the dependence on the resistance R on the temperature T, the following equation applies:

R(T) = R(T0) (1 + a (T - TO)) = RO (1 + K).

wherein K is the temperature dependence of the resistance R of the load and RO is the resistance of the load at a defined temperature TO. In this case, standard values of about 20 degrees, which occur in usual operation, are assumed.

The keying ratio U is determined starting from the resistance RO occurring at normal temperature TO and the factor K. In the case of the normal temperature, the factor K assumes the value zero. The mean current IF arises on appropriate drive control of the switching means 120.

6 The thus determined actual value 11 for the current is compared with the target value ISF at the point 370. Starting from this value, the integrator 350 determines the value K. This takes place, for example, as follows: If the actual value 11 is greater than the target value, the value K is increased by a certain value. This increase takes place until the actual value 11 is equal to the target value ISF If the actual value 11 is smaller than the target value, the value K is reduced by a certain amount. This reduction takes place until the actual value 11 is equal to the target value. The value zero is usually used as an initial value for the integrator.

Subsequently, the value K is limited by physically plausible values in the limiter 340 and is then multiplicatively linked with the value RO at the point 320. The result is then added to the value RO at the point 310. The interlinking at the points 320 and the 310 replicates the above equation.

The keying ratio U is then computed with the thus corrected value R. The thus ascertained value K is then constantly available for computation of the resistance R or for computation of the drive control signal.

The current regulation consists of a preliminary control of the resistance value of the load by means of the standard resistance RO and an adaptation of the temperature-dependent correction factor K. When a change in the target value occurs, this leads directly to a change in the keying ratio U by reason of the preliminary control and thereby to a change in the actual value 11.

Changes in the resistance of the load as a rule take place very slowly. These changes are compensated for by the adaptation of the factor K. Information about the actual value of K is formed from the integral of the regulating difference (11 - ISIF).

By way of the integration time constant of the integrator 350, the speed of adaptation can be set appropriately for the time constant of the resistance-capacitance low-pass filter 141.

The computation of the keying ratio takes place in a fixed time raster. Thus, the computation is performed, for example, every ten milliseconds. The actual value detection has a large time constant.

7 The load 100 reacts very slowly to a changed drive control signal. This means that the regulating section has a very low time constant. Due to the very low drive control frequency, which is preferably in the range of about 200 hertz, a large time constant results for the filtering of the actual value. The actual value is therefore available with a long dead time.

Regulation of such a system with a dynamically good setting member of an actual value detection with a long dead time and a digital regulator with a long scanning time is as a rule highly problematic. This means that the dead time of the filtering procedure determines the regulating speed of the entire regulating loop. By contrast, in the case of the aforedescribed method exemplifying the invention, the particular structure of the preliminary control, which presets the keying ratio in dependence on the preliminary control value RO for the resistance and the desired current value, and the adaptation of the resistance by means of the integrator, allow a very rapid and exact setting of the actual current to be achieved on a change of the target current. The rapid reaction to changes in the target value are ensured by the preliminary control of the resistance value and the regulating accuracy is ensured by the integrator 350, which sets the correction factor K to the required value.

8

Claims

1. A method of controlling a load by a pulsed drive signal with a presettable keying ratio, comprising the steps of presetting the keying ratio in dependence on a resistance value of the load and a target current value, and adapting the resistance value of the load.

2. A method as claimed in claim 1, wherein the step of adapting is carried out in dependence on the difference between the target current value and an actual current value.

3. A method as claimed in claim 1 or claim 2, comprising the step of presetting the resistance value in dependence on a preliminary control value and a correction factor.

4. A method as claimed in claim 3 when appended to claim 2, comprising the step of presetting the correction factor in dependence on the result of a comparison of the target current value with the actual current value.

5. A method as claimed in claim 3 or claim 4, wherein the correction factor is presettable by integrating means.

6. A method as claimed in any one of claims 3 to 5, wherein the step of presetting the resistance value comprises multiplicatively interlinking the preliminary control value and the correction factor and adding the product to the preliminary control value.

7. A method as claimed in claim 2 or claim 4 or any one of claims 3, 5 and 6 when appended to claim 2, comprising the step of subjecting the actual current value to low-pass filtering.

8. A method as claimed in claim 1 and substantially as hereinbefore described with reference to the accompanying drawings.

9. Control means for controlling a load by a pulsed drive signal with a presettable keying ratio, comprising means for presetting the keying ratio in dependence on a resistance value of the load and a target current value, and adapting the resistance value of the load.

9

10. Control means substantially as hereinbefore described with reference to accompanying drawings.